Synthesis from PET to MR and CT modal images using the latent diffusion model

Motivation: Medical images in different modalities (MR\PET\CT) can provide different information, which can help to fully understand a patient's condition and assist physicians in making accurate diagnoses and treatment plans. Goal(s): Using the diffusion model to generate multiple modal images from a single modality. Approach: We propose the conditional latent diffusion model (CLDM) guided by category information to address the challenge of completing the target modal image within the same body. Results: Compared to images generated by GANs, our model produces higher quality images with enhanced capabilities, particularly in capturing intricate details. Impact: Our study offers bright future for diffusion models in the complementary field of medical imaging modalities.